Revitalizing DAC

The 61^st Design Automation Conference is just two months away and as I get closer to retirement, I know there will only be a few remaining for me. I entered the EDA industry in 1980, so have been involved with it for almost 45 years. Over that period, I have only missed a few. It is interesting how the conference has changed over the years.

In the early days, DAC was only a conference, where like-minded people came together to share ideas and to prevent duplicate efforts. It really took off in the early ’80s when EDA companies started to pop up everywhere. I was involved in a little English company that had an RTL simulator called Hilo. It got bought by GenRad, an American tester company, which had been doing fault simulation for many years for the development of functional test patterns. That is how I came to live over here.

Those early days were exciting, with perhaps more theater than product, with wild parties and hotel suites being used for private demonstrations and wooing customers. Conference panels were standing room only, and there was a great diversity in the papers being presented. Each company on the show floor talked about what they were working on, or ideas they had for the future – almost as a litmus test to see which were worth developing further. Going around the show floor was one of the highlights of the event.

As the EDA industry consolidated and innovation slowed, there were still many colorful characters in the industry, and everyone who has been around for a while will remember Joe Costello, then CEO for Cadence, talk about EDA companies feeding from the same dog bowl. This was a sad time for EDA because design starts were going down, and it was projected that less than a handful of companies would ever be able to afford to manufacture with the latest nodes. Researchers lost interest because there were not many projects that could be carved out and suitable for a small team of people. A large amount of research moved to the application of EDA, rather than developments within EDA. The industry became highly risk-averse, meaning that even if there were advancements, they were unlikely to be adopted.

Along the way, there has been plenty of drama. Large EDA companies felt they no longer got value from the show and threatened to pull out. They already had their own private conferences where they had full control over the customers at those events. Some even did disappear for a year or so before returning. DAC always has been a good place for startups, being one of the few places they can get in front of a large number of potential customers who might not have heard about them. When the Internet bubble burst, many companies put in place travel bans to cut costs, and conferences were the first to go. A few managers continued to attend, but engineers became a rare sight. They have never really returned.

Today, the parties are replaced by somewhat sophisticated receptions. Most company parties have gone. There is no official DAC party. But perhaps, to me, the saddest changes are on show floor. Booths are nothing more than foyers to private presentation rooms, and while there may be some demos being shown, they contain little new information. It is not clear who they are really for.

At the same time, the conference is coming back to life. The CHIPS Act has brought attention to the industry, not just semiconductors, but EDA itself. The slowdown in Moore’s Law has created a plethora of research opportunities because improvements in performance or power can no longer be dealt with by moving to the next node. Architecture, design, software, AI, safety, security – these all have become fertile grounds for innovation and it is reflected in the conference program.

But what about the show floor? What will it take to revitalize that? It is clear that the program committee has tried very hard over the past few years to bring in new companies, new events, add excitement. They are certainly working, but the booths are not responding. A few companies have tried holding design or verification challenges – at least those make it more interactive. The autonomous driving challenge was great – too bad it did not become a feature. I loved trying to get younger people involved – more of that please. It is time for the big EDA companies to step it and innovate how they get people excited about the industry, their products, make people want to come to the show floor.

What would you like to see happen to the show floor? Perhaps we need to start giving them some ideas.